☰

Help center › Data and Interfaces

The loading center

Load reference data, holdings and market data from a file or a folder: column mapping, deterministic validation, a diff before anything changes, apply and undo.

At a glance

Where
System › Data and Interfaces › Loading center (/app/data)
Who
Users with load data: Operations and Admins. Only an Admin loads users and roles. Roles and permissions

Try it in the demo A private sandbox firm, no sign-up. Simulated data only.

On this page

Every file is staged first: you see what is new, changed, unchanged and rejected before anything changes, and an applied batch can be undone.

The loading center
Upload a file, or a whole folder; data quality and the data sets below.

Load a file

  1. What the file holds: pick the data set (calendars, blackout windows, issuers, security master, classifications, security lists, accounts, account groups, model portfolios, benchmarks, brokers, positions, tax lots, cash balances, end-of-day or intraday prices, FX rates, average daily volume, corporate actions, risk factors).
  2. File: .csv, .txt, .xlsx or Vestoris .xml; .gz or .zip; up to 50 MB.
  3. Mode: Add and update (delta), or Complete file (full): records missing from the file are inactivated.
  4. Business date (optional), then Check the file.
  5. Column mapping: columns are matched to fields by name; the AI suggests the rest. You see the mapping and can change it; it is remembered for files with the same columns. Vestoris XML needs no mapping.
  6. Validation is deterministic: ISIN, CUSIP, SEDOL and FIGI check digits, references to accounts and securities that exist, duplicates in the file, dates and numbers. Each rejected row has a code (R101, R104…) and the error file lists them. "Explain in plain English" asks the AI to word the reasons; the codes stay the same.
  7. The diff: new, changed (old → new), unchanged, rejected, deleted or inactivated.
  8. Apply: one transaction; every change is recorded. Or Discard.
A batch and its diff
A batch: its file, counts, the diff row by row, Undo and the original file.

Undo

An applied batch can be undone as a whole from its page (Load history). If something depends on it since (for example a trade on loaded lots) the undo is refused and the reason is shown.

Upload a folder

Send a whole morning's files at once: Choose a folder…, drop a folder on the page, or choose a .zip of it (on a phone or tablet, zip it first). Up to 200 files, 50 MB each, 250 MB in all. Each file is recognised by its XML namespace, a feed's file name, or columns you have loaded before; when only the AI can guess, it suggests and you confirm. On the review screen change any file's data set or exclude it, then Check all files: they are checked in start-of-day order, each as its own batch with its diff. Apply all clean files applies them in order (a file whose dependency has errors waits), and Undo the folder undoes them newest first. Details: Uploading a folder.

Uploaded folders
Each folder with where it came from, its files, business date and status.

Sample files

The demo has samples for the tour: a vendor security master with mistakes in it, an end-of-day price file with one spike, and a sample start-of-day folder (.zip).

The record

Every batch is in the event log with who loaded it, when, the file's SHA-256 and the counts. The original file is stored encrypted for its retention period and can be downloaded from the batch.

  • Load history Every batch: who loaded it and when, the file's SHA-256, the counts and its status; open one for the diff, the rejected rows, undo and the original file.
  • Templates and field dictionary Each data set's fields, what they accept and which column names are recognised; CSV templates and sample XML.
  • Feeds and start of day Feeds are data sets that arrive on a schedule (drop box, HTTPS push or pull); each file becomes a batch, and the start-of-day chain applies them in order.
  • Data quality Deterministic checks over the loaded data: stale or jumping prices, missing classifications or identifiers, negative cash and more. Issues resolve themselves when the data is fixed.
  • History and as-of What any loaded record looked like on any date, and every version with the batch that made it.
  • User-defined fields Ten fields of your own on each reference table: define them, validate them, and use them on forms, the blotter, in loads, compliance and workflow rules.

Do it with the API

Every action on this screen can also be done by your own systems, under the same permissions. See Authentication to get a token.

Paper trading only: orders and fills in Vestoris are simulated. Nothing here is investment advice.